BIND segway -> python -> first-class ports

[John-Mark Gurney]

Alfred Perlstein wrote this message on Mon, Dec 09, 2013 at 18:49 -0800:
> 
> On 12/9/13, 6:36 PM, John-Mark Gurney wrote:
> >Alfred Perlstein wrote this message on Mon, Dec 09, 2013 at 18:26 -0800:
> >>On 12/9/13, 2:11 PM, Daniel O'Connor wrote:
> >>>On 9 Dec 2013, at 22:21, Alfred Perlstein <bright at mu.org> wrote:
> >>>>>On Dec 8, 2013, at 9:13 PM, "Daniel O'Connor" <doconnor at gsoft.com.au>
> >>>>>wrote:
> >>>>>
> >>>>>(I don't care if it's JSON, XML or smoke signals just so long as it
> >>>>>isn't hideously slow).
> >>>>>
> >>>>You have described exactly what we are trying to accomplish. Thank you!
> >>>OK, great, can you have an implementation by COB Friday? :)
> >>>
> >>>Although seriously, I do wonder what the best way to implement it would 
> >>>be.
> >>>
> >>>I can imagine modifying /bin/sh to have such a feature would be somewhat
> >>>controversial..
> >>>
> >>A few interesting targets would be sysctl, ifconfig, gpart, disklabel,
> >>... any others?
> >gpart info is already in XML from sysctl kern.geom.confxml...
> >
> >And disklabel has been retired and replaced by gpart...
> >
> XML doesn't work for everyone, and as far as disklabel, wasn't aware it 

hence the thread, but my point is some of this already exists...

> was toast in HEAD, it's still around in 9.2 and we've been doing most of 
> our development there so oops!

well, disklabel aka bsdlabel (to be platform neutral) is still around,
but I do not believe it works on disks where one of the partitions are
open, while gpart will:
bsdlabel: cannot open provider /dev/ada0s2 for writing label: Permission denied
bsdlabel: Try to use gpart(8).

So, you should be using gpart to be future compatible, and not limit
yourself to one type of partitioning scheme..

-- 
  John-Mark Gurney				Voice: +1 415 225 5579

     "All that I will do, has been done, All that I have, has not."